Mice

Generation and genotyping of Piezo1fl/fl and Piezo2fl/fl mice have been described previously.10,51Col2a1Cre mice52 were purchased from the Jackson Laboratory (Bar Harbor, ME, USA; #003554). We crossed Piezo1fl/fl, Piezo2fl/fl or both with Col2a1Cre mice and analyzed animals from the resulting mouse strain (STOCK Piezo1tm1c(KOMP)WtsiPiezo2tm2.2Apat Tg(Col2a1-cre)1Bhr/Uke). To exclude a possible influence of genetic background, all analyses were performed with the corresponding Cre-negative littermates. Mice were housed in a specific pathogen-free environment with a 12 h light-dark cycle, 45% to 65% relative humidity, and 20 °C to 24 °C ambient temperature in open or individually ventilated cages with wood chip bedding and nesting material in groups of no more than six animals. Mice had ad libitum access to tap water and standard rodent food (Altromin Spezialfutter GmbH & Co. KG, Lage, Germany;1328 P). Animal breeding and all experimental procedures with defined humane endpoints were performed with approval from the animal care committees of the University Medical Center Hamburg-Eppendorf and the Hamburg Ministry of Justice and Consumer Protection (N033/2022, N031/2021). No preregistration was performed.

Anterior cruciate ligament transection (ACLT)

The anterior cruciate ligament transection (ACLT) procedure was performed on the right knee joint of 12 weeks old female Piezo1Col2a1cre (n = 8) and Piezo2Col2a1cre (n = 9) mice as well as Cre-negative littermates (n = 10) to induce OA, as previously described.53 To minimize confounders, batch surgeries were performed with the inclusion of different genotypes per batch. Briefly, mice were anesthetized using isoflurane (4% isoflurane for induction and 1%–2% for maintenance), and the right knee joint was accessed via a medial parapatellar incision. Using a scalpel with a pointed blade, the anterior cruciate ligament of the knee joint was transected under vision using 2.5x magnification glasses. Complete transection was confirmed by a positive anterior drawer test, indicating successful ACLT. Prior to surgery, mice were administered a prophylactic dose of 150 mg/kg clindamycin subcutaneously to reduce the risk of infection. To manage post-operative pain, mice were given 0.1 mg/kg buprenorphine subcutaneously. Both treatments were administered in a sterile manner to minimize the risk of infection and contamination. Additionally, 2 g/L metamizole was added to the drinking water for three days pre- and post-surgery. The contralateral (left) knee joints served as a non-operated control to account for any systemic changes or individual differences that may affect the study outcomes. Non-operated control joints were not subjected to any surgical intervention and were processed identically to ACLT joints. After surgery, the welfare of mice was assessed daily based on overall appearance and body weight. One animal was excluded from the analysis due to a post-surgery complication. Animals were sacrificed 8 weeks post-surgery via CO2 intoxication. After dissection, knee joints were aligned at an angle of 90° and skeletons were fixed in 3.7% PBS buffered formalin for 48 h. Contact radiography was performed for all animals after fixation. All endpoint measurements described below were performed in a blinded fashion.

Micro-computed tomography (µCT)

To determine the bone microstructure in knee joints, micro-computed tomography (µCT) was performed using a µCT 40 cone-beam system (SCANCO Medical AG, Brüttisellen, Switzerland) with a voxel size of 10 µm. The X-ray tube was operated at 55 kVp and 145 µA intensity. To investigate OA-related changes in both the ACLT and non-operated control knees, the subchondral bone and osteophytes of the tibia were evaluated using µCT. Trabecular bone microarchitecture parameters were analyzed in the subchondral region extending until the growth plate. For determination of osteophyte parameters manual segmentation was applied.

Due to frequent rib fractures in 7-day-old mice, rib morphology was not assessable by standard µCT. To determine reasons for early fracturing, the trabecular and bone microstructure of ribs in newborn (P0) and 6-week-old Piezo1Col2a1Cre mice was analyzed. Ribs were scanned using a desktop µCT system (Skyscan 1272, Bruker, Kontich, Belgium) at a resolution of 1.4 µm and 3.4 µm voxel size for P0 and 6-week-old mice, respectively. Per sample, the values obtained from four ribs were averaged. The X-ray tube was operated at a voltage of 50 kVp and an intensity of 200 µA. For quantification, the trabecular and cortical microstructure was evaluated separately in a region of interest from 0.5 mm to 8 mm distal (Fig. 2d) of the growth plate. Specifically, the cortical thickness (Ct.Th in µm) and the bone volume fraction (BV/TV in %) were determined.

Undecalcified histology and histomorphometry

Bone samples were dehydrated in increasing concentrations of ethanol and embedded in methylmethacrylate for undecalcified histology, using previously described protocols.25 For knee joints, 4 µm thick coronal plane sections, were obtained and stained using Safranin-O, trichrome Goldner, toluidine blue, hematoxylin-eosin (HE), and Movat Pentachrome staining as described before.54,55,56 The severity of OA-related changes in all four compartments of the knee joint was assessed using the Osteoarthritis Research Society International (OARSI) scoring system for mice, with scores ranging from 0 (normal joint) to 24 (severe OA).57 Each compartment of the joint was evaluated separately and then summed to obtain an overall score for each mouse. The assessment of the osteochondral unit included cartilage thickness, calcified cartilage thickness and subchondral bone plate thickness, which were assessed for each compartment of the joint separately. For each parameter, the values of the compartments were averaged to provide a single value per knee joint. The histopathological synovitis score was used to assess inflammatory changes in the knee joints.58 Briefly, the enlargement of synovial lining cells and their density were evaluated individually. The formation of osteophytes was evaluated using a scoring system based on osteophyte size and maturity, as previously described.59 For histomorphometric analysis of knees and lumbar spine, sagittal plane sections were obtained and stained with toluidine blue, Safranin-O, and the von Kossa/van Giesson procedure as described previously.60 Trabecular parameters (Bone volume fraction (BV/TV in %), trabecular number (Tb.N in mm−1)) were quantified using the BioQuant Osteo Software (BIOQUANT Image Analysis Corp., Nashville, TN, USA) in accordance with the ASBMR guidelines.61 Cellular histomorphometric parameters [Number of osteoblasts per bone perimeter (N.Ob/B.Pm in mm−1), number of osteoclasts per bone perimeter (N.Oc/B.Pm in mm−1), osteoblast surface per bone surface (Ob.S/B.S in %) and osteoclast surface per bone surface (Oc.S/B.S in %)] were quantified in undecalcified toluidine blue stained sections of knees after ACLT surgery and lumbar spine of 6-week-old mice using the Osteomeasure system (OsteoMetrics, Inc., Atlanta, GA, USA) in accordance with the ASBMR guidelines.61 Growth plate thickness and chondrocyte number were measured on toluidine-stained sections using the Osteomeasure system (OsteoMetrics).

Serum levels of bone resorption markers

The levels of bone resorption markers in the serum of ACLT mice after final exsanguination were measured using the respective enzyme-linked immunosorbent assays (ELISA) according to the manufacturer’s instructions. Levels of receptor activator of nuclear factor kappa-Β ligand (Rankl) were measured using the respective Quantakine Elisa Kit (R&D Systems, Minneapolis, MN, USA; MTR00). The concentrations of C-terminal telopeptide of type I collagen (CTX-I) were determined using the RatLaps kit (Immunodiagnostic Systems Holdings Ltd., Boldon, UK; AC-06F1).

Cell culture

Murine chondrogenic ATDC5 cells (Sigma-Aldrich; 99072806) were cultured in DMEM:F12 (Gibco, Billings, MT, USA; 11320-033) with 5% FCS (Gibco; 10437-028) and 100 U/mL penicillin and streptomycin (Gibco; 15140122) at 37 °C, 5% CO2 and 95% relative humidity.

Primary murine chondrocytes were isolated from rib cage cartilage of 10-day-old mice by collagenase (Sigma-Aldrich; C9891) digestion in DMEM:F12 and cultured in DMEM:F12 with 10% FCS and 50 μg/mL ascorbic acid at 37 °C, 5% CO2 5% O2 and 95% relative humidity.

Treatment with Yoda1 and Verteporfin

ATDC5 cells were seeded at a density of 4.3 × 104/cm2. Primary murine chondrocytes were seeded at a density of 6.3 × 104/cm2 and cultured for four days until confluent. Cells were serum-starved overnight and subsequently stimulated for 6 h in serum-free growth medium with the Piezo1 agonist Yoda1 (Tocris Bioscience, Bristol, UK; 5586, 5 μmol/L) and/or the YAP inhibitor Verteporfin (Selleck Chemicals LLC, Houston TX; S1786). Control cells were treated with an equal amount of DMSO.

In vitro RNA expression and transcriptome analysis

RNA from cultured cells was isolated and subjected to DNase digestion using the NucleoSpin RNA kit (Macherey-Nagel GmbH & Co. KG, Düren, Germany) according to the manufacturer’s instructions. Concentration and quality of RNA were measured using a NanoDrop ND-1000 system (Thermo Fisher Scientific, Waltham, MA, USA) and the TapeStation 2200 system (Agilent Technologies, Santa Clara, CA, USA). Genome-wide expression analysis was performed using the Clariom D system (Thermo Fisher Scientific) as described previously.25 In brief, the Clariom D assay system (Thermo Fisher Scientific) was used with 100 ng of total RNA as starting material to generate 2nd-cycle single-stranded complementary DNA (ss-cDNA). 5.5 μg of ss-cDNA were used for gene chip hybridization (Clariom D, mouse, Thermo Fisher Scientific). Gene chips were washed and stained using the Affymetrix Fluidics Station 450 and scanned with the Affymetrix Gene Chip Scanner 7 G (both Affymetrix, Santa Clara, CA, USA). Data were analyzed in the Transcriptome Analysis Console software (TAC 4.0; Thermo Fisher Scientific) using default analysis settings (version1) and Gene + Exon-signal space transformation-robust multiarray analysis (SST-RMA) as summarization. Full gene expression datasets have been deposited at the GEO database (National Center for Biotechnology Information [NCBI], Bethesda, MD, USA; https://www.ncbi.nlm.nih.gov/geo/) with accession number GSE230071.

For qRT-PCR expression analysis, 500 ng of total RNA was reverse transcribed using Verso cDNA Synthesis Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions using oligo-dT primers. Quantitative expression analysis was performed using a StepOnePlus system and predesigned TaqMan gene expression assays (Thermo Fisher Scientific). Gapdh expression was used as the reference housekeeping gene. The following assays were used: Ptgs2 (Mm00478374_m1), Ccn2 (Mm01192933_g1), Sox9 (Mm00448840_m1), Dusp1 (Mm00457274_g1), Ereg (Mm00514794_m1), Il1rl1 (Mm00516117_m1), Dusp6 (Mm00518185_m1), Ier3 (Mm00519290_g1), Piezo1 (Mm01241549_m1), Piezo2 (Mm01265861_m1), Col2a1 (Mm00491889_m1), Acan (Mm00545794_m1). Data analysis was performed according to the delta-delta comparative threshold cycle (2-ΔΔCT) method, and results are shown as fold-change expression values relative to controls.

Transdifferentiation assay with Ccn2/Ctgf

The transdifferentiation assay was performed as described before.41 Briefly, ATDC5 cells were cultured in chondrogenic differentiation medium for 7 days and thereafter cultured in osteogenic differentiation medium containing 50 ng/mL recombinant human Ctgf (PeproTech Inc., Rocky Hill, NJ, USA). Cells were harvested after 24 and 48 h of osteogenic differentiation, respectively. RNA from cell lysates was isolated using the RNeasy Mini Kit and cDNA was generated using Omniscript Reverse Transcriptase (both Qiagen). qRT-PCR expression analysis was performed using Platinum™ SYBR™ Green qPCR SuperMix-UDG and ROX™ Reference Dye (both Thermo Fisher Scientific). The primer sequences of the analyzed genes are shown in Table S1. B2m expression was used as the reference housekeeping gene. Data analysis was performed according to the delta-delta comparative threshold cycle (2-ΔΔCT) method, and results are shown as fold-change expression values relative to controls.

In vivo RNA expression analysis

For expression analysis after OA induction, twelve-week-old C57BL/6 J wildtype mice were subjected to ACLT as described above, or a sham operation in which only the medial parapatellar incision was performed. Whole knees were carefully isolated from 20-weeks-old C57BL/6 J mice at 8 weeks after ACLT or sham surgery. Tissue was snap frozen in liquid nitrogen, crushed with mortar and pestle followed by homogenization in TriFast (Peqlab, Erlangen, Germany; 30–2010) using an UltraTurrax T25 (IKA, Staufen, Germany). RNA isolation, cDNA synthesis and qRT-PCR were performed as described above for cultured cells.

For expression analysis of growth plate and bone samples, tibia and femur were isolated from 12-day-old mice. Distal femoral and proximal tibial growth plates and cortical bone samples were isolated using ophthalmic scalpels under a stereo microscope. Tissue was snap-frozen in liquid nitrogen and homogenized using a micropestle. RNA was isolated using a NucleoSpin RNA XS kit (Macherey Nagel GmbH & Co. KG; 740990.50). CDNA synthesis and qRT-PCR were performed as described above.

Human specimens

Eight femoral heads were collected from patients with primary OA undergoing total hip arthroplasty, comprising of 4 women and 4 men. Control specimens were also obtained from eight individuals (5 women and 3 men) during autopsy. Osteophytes were identified and excised from the margins of the femoral head specimens. The identification of osteophytes was done macroscopically and based on the characteristic morphological features of osteophytes, including their firmness and their localized growth at the joint margins. An excision margin of 1 mm was maintained around the osteophytes to ensure that articular cartilage and normal bone were not included in the excised material. All samples were fixed in 3.7% formaldehyde within 24 h of death or resection. To prepare histological sections, the samples were cut along the coronal plane through the fovea capitis femoris using a diamond-coated saw, with the medial part of the femoral head selected for analysis. Both decalcified and undecalcified histology specimens were prepared. For undecalcified histology, the specimens were embedded in methylmethacrylate, cut into 4 µm sections, and stained with toluidine blue, von Kossa, and Safranin-O. The fractions of the different components of the osteophyte, specifically the osteophyte rim covered with cartilage, zones of endochondral ossification, and fibrotic zones, were manually traced and quantified using the Osteomeasure system (OsteoMetrics, Inc., Atlanta, GA, USA).

Immunohistochemistry

For immunohistochemistry of human samples, femoral head and osteophyte specimens were prepared as described above and were decalcified in an EDTA-based solution (Usedecalc, Medite, Orlando, USA) for 2 weeks at 20 °C. Decalcified samples were dehydrated in increasing concentrations of ethanol and embedded in paraffin. Coronal sections (4 μm) were prepared from the decalcified specimens. Heat-mediated antigen retrieval was performed at 60 °C in a water bath overnight. Endogenous peroxidase was blocked with Bloxall (Vector Laboratories Newark, CA, USA; SP-6000-100) and unspecific binding was blocked with 5% normal goat serum for 1 h. Sections were incubated overnight with the following primary antibodies in 2.5% goat serum: Rabbit anti-Ccn2/Ctgf (1:100, Novus Biologicals, Littleton, CO, USA; nb100-724), rabbit anti-Ptgs2 (1:100, Proteintech, San Diego, CA, USA; 12375-1-AP) or rabbit anti-Piezo1 antibody (1:50, Proteintech; 15939-1-AP) overnight at 4 °C. After washing, sections were incubated with Dako EnVision Flex HRP (Agilent Technologies; SM802) as secondary antibody for 30 min. Staining was developed by incubation with the chromogenic substrate 3,3′-Diaminobenzidine (DAB Substrate Kit, Vector Laboratories; SK-4100). Slides were counterstained with Mayer’s Hematoxylin (Sigma-Aldrich, St. Louis, MO, USA; MHS16), blued under running tap water, dehydrated and mounted with DPX mountant (Sigma-Aldrich, 44581). Semi-quantitative scoring of positively stained cells, with a grading system ranging from 0 to 4, where no staining was graded as 0, and strong staining of all cells was graded as 4, was performed independently by three blinded observers. The quantification was performed for the articular cartilage, subchondral bone plate, and trabecular bone of the femoral head, as well as for cartilage, zones of endochondral ossification, bone and fibrosis within the osteophytes. The scores of the observers were then averaged for each compartment.

For immunohistochemistry of mouse lumbar spine sections, lumbar spines from 2-week-old mice were decalcified in 20% EDTA for one week, and sagittal plane paraffin sections were prepared as described above. Deparaffinized sections were treated with pepsin for 60 min at 37 °C (Agilent, SantaClara, CA, USA; S3002), followed by hyaluronidase digestion (0,5%; Sigma-Aldrich; H3506) for 20 min at 37 °C. Peroxidase and serum blocking were performed as described above. Sections were then incubated with the following primary antibodies in 2.5% goat serum: Rabbit anti-collagen type I (1:100, Kerafast, Boston, NE, USA; ENH018-FP), rabbit anti-collagen type II (1:100, Abcam, Cambridge, UK; ab34712) for 1 h at room temperature. Secondary antibody incubation and further processing were performed as described above.

TRAP staining

For the staining of tartrate-resistant acid phosphatase (TRAP), lumbar spine paraffin sections from two-week-old mice were incubated in a substrate solution (40 mmol/L sodium acetate, 10 mmol/L sodium tartrate, 1.6 mmol/L fast red violet, 700 mmol/L naphthol, pH 5; all chemicals Sigma-Aldrich) for 90 min and counterstained with Mayer’s hematoxylin (Sigma-Aldrich). Quantification was performed using the Osteomeasure (OsteoMetrics, Inc.) system as follows: The hypertrophic zone and underlying primary spongiosa were marked as “bone” and TRAP-positive cells were marked as “osteoclasts”. The parameter number of osteoclasts per bone perimeter (N.Oc/B.Pm) was derived, which represents the number of TRAP-positive cells per tissue perimeter (TRAP+ cells/T.Pm in mm−1).

Statistical analysis

Data are shown as box plots with dots representing individual values and indicating the group size. Statistical analysis was performed using SPSS (version 29, IBM, Armonk, USA) and GraphPad Prism (version 9, GraphPad Software, La Jolla, USA) for visualization. For comparison of more than two groups one-way ANOVA (one variable) or two-way ANOVA (two variables) with Tukey post-hoc test was performed. Comparisons between two groups were performed with the Student’s t test. All statistical tests were two-sided and a significance threshold of 0.05 was chosen.